Most of the allogeneic transplants performed at our and other institutions involve matched donors. Despite HLA-matching, graft-versus-host-disease (GVHD) still occurs in 50% of these cases. Attempts have been made at our own institution to combat GVHD by removing T cells from the donor graft prior to transplant. This T cell depletion (TCD), although reducing the severity of GVHD, has increased the frequency of graft rejection. In this proposal, we intend to utilize our expertise in murine transplant models to construct a murine model of TCD-induced graft rejection across minor histocompatability barriers. We will use these models to evaluate immunotoxins as potential agents for overcoming graft rejection. Immunotoxins (IT) are monoclonal antibodies (moab) linked to potent catalytic toxins such as ricin toxin A chain (RTA) and are designed to deliver a lethal hit to specific antigen expressing target cells. These agents are currently under clinical evaluation for GVHD therapy at our and other institutions. Our specific aims are to test the hypothesis that RTA IT administered in vivo will promote engraftment across the minor histocompatibility barrier. Since in vivo attempts to define the role of immune cells in the recognition of minor histocompatibility antigens have been hampered by the unavailability of agents that can reliably and selectively deplete target cell populations, we intend to use our experience with RTA IT to define the cell populations involved in the engraftment network. Various RTA IT will be studied for in vitro selectivity and potency and then studied in vivo. We will determine optimal conjugation conditions, dosage, route, and schedule for each IT. Efficacy of engraftment promotion will be measured in both short-term and long-term engraftment systems. In vivo depletion of target cells will be monitored using fluorochrome labeled moab and flow cytometry and function. We also intend to test the hypothesis that the presence of the translocation-enhancing regions of toxin makes better IT. Many catalytic toxins such as ricin and pseudomonas exotoxin A (PE) have an enhancing region that promotes translocation of the A chain into target cells. This region is separate from the native binding site that renders these toxins dangerous to man. The state of the art now permits us to produce recombinant toxin devoid of the native binding region, but containing the toxic A chain and the translocation enhancing region. In this proposal we will study such a toxin, called PE40, as an IT in our BMT model, PE40IT will be studied and compared to RTA IT.